Does bimodally seasonal rainfall promote species-diversity in large mammals, worldwide?

@paradoxornithidae @tonyrebelo @jeremygilmore

Everyone knows of the diversity of large mammals in Kenya (

However, how many naturalists understand the ecological reasons for this diversity?

One of the environmental factors distinguishing Kenya is that, in parts of this country, there are two rainy seasons each year (

This bimodality extends from Kenya to southern Uganda ( and northernmost Tanzania, as well as broadly to the Horn of Africa (

Bimodally seasonal rainfall is likely to boost the reliability of food through the annual cycle, for the indigenous ungulates.

In turn, this reliability hypothetically means that forms can specialise on particular foods, producing narrow niches, and allowing various closely-related species to coexist to a degree that might be impossible otherwise.

The result, if this principle is valid, would be faunas and communities of large mammals that - other factors being equal - are richer in species than those in climates with unimodally seasonal rainfall.

There at least 44 spp. of ungulates indigenous to Kenya (

The gerenuk (Litocranius walleri, seems to epitomise adaptation to bimodal seasonal rainfall.

This species of gazelle, occurring in Kenya and adjacent countries, is possibly the most specialised ungulate on Earth. It relies on the shoots of spinescent acacias, and has an accordingly aberrant body-form and posture.

Once one realises that Kenya has an unusual climate w.r.t. bimodally seasonal rainfall, a search-image arises for other regions on Earth that also show - albeit to a lesser degree - a similar correlation between climate and fauna.

Please see

This global map shows the regions as

  • reddish for rainfall regimes that have bimodally seasonal rainfall, and
  • whitish for adjacent regimes with some degree of bimodality, transitional to unimodally seasonal rainfall.

The more strongly unimodal the rainfall, the more intensely blue the portrayal is, in this map. For example, most of south-central and West Africa, plus the Sahel (, has rainfall simply concentrated in summer (e.g.

The effects of bimodally seasonal rainfall on the reliability of food for large mammals are likely to be outweighed/irrelevant where the climate is

  • arid (e.g. see the areas marked reddish in Egypt, and near Mecca in Saudi Arabia),
  • frigid (e.g. Greenland), or
  • pluvial (e.g. Congo/Cameroon/Gabon, northwestern Colombia, southern British Columbia, and northern Kalimantan on the island of Borneo).

Therefore, let us focus on those regions with mesic rainfall (250-1000 mm mean annual precipitation) and moderate temperatures.

The following regions, featuring reddish on the map, are noteworthy:

Of the regions mentioned above, the ones that I find climatically most surprising are those in Pakistan and Spain.

I have yet to understand how bimodality has arisen here, and I have yet to assess the indigenous fauna, given that these regions have long been intensely affected by human activity.

However, I can comment on the treeless grassland biome, found on all vegetated continents.

This biome is extremely inconsistent in the incidence and diversity of large mammals, as follows:

In the Highveld ( of South Africa, as many as 18 species of ungulates occur sympatrically (please see Table 1 in

This contrasts with South America: in the pampas of Buenos Aires province of Argentina, only two species occur.

North America and Eurasia have intermediate numbers.

Finally, in Australia - where there are no indigenous ungulates - macropod marsupials ( occur, but surprisingly few species penetrate the biome in question (

Is this variation explained by the seasonal distribution of rainfall?

The answer is obviously 'no'.

This is because the various regions of treeless grassland vary greatly in the seasonal distribution of rainfall, but with no apparent correlation with the fauna.

The region of treeless grassland in Eurasia has rainfall that is somewhat bimodally seasonal, which differs from North America. Both have similarly modest diversities of ungulates, with the caveat that in Eurasia the original fauna of the Holocene is not fully known. after thousands of years of human influence.
The puzzle is that the region of maximal diversity, viz. the Highveld, has unimodally seasonal rainfall ( By contrast, that of minimal diversity, viz. the pampas, has rainfall that is not strongly unimodal (

Turning now to biomes other than treeless grassland:

None of the other regions (Sri Lanka, Arizona, Spain, Pakistan) has a particularly diverse fauna of large mammals.

In the case of Sri Lanka, the situation is complicated by the fact that this has been an island since the start of the Holocene. Insularity can be expected to limit diversity.

The southernmost part of the Indian subcontinent, just north of Sri Lanka, has a tendency towards bimodally seasonal rainfall (

This region does have a diverse fauna of large mammals by Asian standards. For example, the indigenous artiodactyls of the Holocene are Sus scrofa (, Axis axis (, Rusa unicolor (, Muntiacus vaginalis (, Nilgiritragus hylocrius (, Antilope cervicapra (, and Moschiola indica (, plus possibly the wild ancestor of Bubalus bubalis in former times.

However, this fauna is not significantly more diverse than that farther north in India, where rainfall is unimodally seasonal.

I conclude, overall, that the pattern observed in Africa does not - for whichever reasons - seem to apply to other continents.

Posted on August 20, 2023 12:30 AM by milewski milewski


The equatorial zone in Africa has bimodally seasonal rainfall, even in the rainforest biome of Equatorial Guinea (, Gabon (, Cameroon (, Congo (, and the Democratic Republic of the Congo (

Surprisingly, this is not replicated in South America.

Here, the equatorial rainforest has bimodally seasonal rainfall only in Colombia, the bimodality failing to extend along the equator to the same biome in Brazil, Peru, Ecuador, Venezuela, and Guyana ( and

This continental difference extends to the Horn of Africa relative to easternmost Brazil (including the caatinga, The former includes the most bimodally seasonal rainfall on Earth (shown in reddish brown on the map), whereas the latter lacks the bimodal pattern altogether.

The eastern United States and China are climatically comparable, despite a considerable disparity in latitude.

An intriguing difference is that only in the former does bimodally seasonal rainfall occur (see page 24 in

Posted by milewski 11 months ago

Bimodally seasonal rainfall occurs in the state of Arkansa in the USA.

The following ( shows that there is a peak in mean monthly rainfall in April (spring), and another in December (winter.).

The vegetation is deciduous forest.

Posted by milewski 11 months ago

Add a Comment

Sign In or Sign Up to add comments